Modeling and estimation of the bullwhip effect in multi-stage supply chains: An implementation study in the white goods industry

The concept of supply chain management continues to gain importance in today's competitive conditions. Supply chain management is crucial to ensuring sustainability of companies and maintaining their presence in the competitive market. There are two phenomena in supply chain management are very important which are material flow and information flow. Managing flows is too hard because there is a wide network from supplier to end customer. In supply chain, order variance can occur which is called bullwhip effect. Bullwhip effect causes increasing of production, inventory, labor and transportation cost. It also causes reducing the ability of companies to sustain their competitive market assets. Predicting the bullwhip effect is very important in terms of minimizing the bullwhip effect. In this thesis, the bullwhip effect was tried to be estimated in a 3-stage supply chain. In the first section of this thesis, supply chain management is explained. Supply chain elements and the supply chain flows are explained. And then the bullwhip effect that emerged in the supply chain was defined. The causes of the bullwhip effect and the consequences of the bullwhip effect are defined. Then 'Artificial Neural Networks' and 'Neuro Fuzzy' methods are defined which used on application. In the application section, which is the last stage of the thesis, the most suitable estimation method was tried to determine for predicting bullwhip effect by using company's data. As a result , 'Artificial Neural Networks' method has been shown to give better results than "Neuro Fuzzy". Keywords: Supply Chain Management, Bullwhip Effect, Artificial Neural Networks, Neuro Fuzzy